viewclust.slurm package

Submodules

viewclust.slurm.mem_info module

viewclust.slurm.mem_info.mem_info(d_from, account, fig_out='', debugging=False)[source]

Script for profiling the memory usage of an account via sacct.

DEPRECATION WARNING.

Always outputs various statistical measures to stdout, but can also plot information.

Parameters:
  • d_from (date str) – Beginning of the query period, e.g. ‘2019-04-01T00:00:00’.
  • account (str) – Account to query via sacct, e.g. ‘def-tk11br_cpu’
  • fig_out (str, optional) – Writes the generated figure to file as the given name. If empty, skips writing. Defaults to empty.
  • debugging (boolean, optional) – Boolean for reporting progress to stdout. Default False.

viewclust.slurm.sacct_jobs module

viewclust.slurm.sacct_jobs.sacct_jobs(account_query, d_from, d_to='', debugging=False, serialize_frame='', slurm_names=False)[source]

Ingest job record information from slurm via sacct and return DataFrame.

Parameters:
  • account_query (str) – String query to be sent to sacct via -A flag.
  • d_from (date str) – Beginning of the query period, e.g. ‘2019-04-01T00:00:00
  • debugging (boolean, optional) – Boolean for reporting progress to stdout. Default False.
  • sacct_file (str, optional) – Loads a raw query from file. If empty, query is rerun. Defaults to the empty string.
  • serialize_frame (str, optional) – Pickle the resulting DataFrame. If empty, pickling is skipped. Defaults to the empty string.
  • slurm_names (str, optional) – Keep slurm’s sacct column names instead of shorthands. Defaults to False.
Returns:

Returns a standard pandas DataFrame, or an empty dataframe if no jobs are found.
Return type:

DataFrame

Module contents